Galaxy Q2 2026: Crypto Losses Persist, Data Centers Ramp Still Invisible

What happened

Galaxy reported a Q2 2026 net loss of $(85) million and adjusted EBITDA of $(77) million, with digital asset price depreciation again cited as the primary driver. This marks a sequential improvement from Q1’s $(216) million loss but underscores that the core crypto platform remains a significant earnings drag. Despite delivering the first Helios data hall in April and guiding for a Data Centers revenue ramp, adjusted gross profit of just $43 million and deeply negative adjusted EBITDA indicate that the segment’s contribution was minimal and failed to offset crypto headwinds. Total equity of $2.7 billion and $2.5 billion in cash and stablecoin holdings provide a liquidity buffer, but the company still requires additional financing to complete Helios conversion, keeping leverage and dilution risks alive. The quarter offers little evidence that the AI data center pivot is yet altering the earnings profile, leaving the stock heavily tethered to crypto sentiment and execution milestones.

Implication

Galaxy’s Q2 results reinforce that the transition to a contract-backed AI infrastructure model is slow and still overrun by crypto-cycle volatility. While Phase I of Helios is progressing, the negligible Data Centers revenue and negative segment profitability mean the company remains a crypto beta play with an expensive infrastructure option attached. The balance sheet is substantial, but the ongoing need for external financing to fund Helios construction adds dilution and covenant risk, especially if digital asset prices weaken further. The WAIT rating from the prior analysis remains appropriate; an entry is only justified once a quarter of material Data Centers revenue and positive adjusted EBITDA is reported. Until then, the stock’s valuation will remain hostage to crypto headlines and any Helios delivery slip could trigger a sharp de-rating.

Thesis delta

The Q2 results indicate that the anticipated Helios-driven earnings inflection has not yet materialized, delaying the expected shift from a crypto-cycle exposed balance sheet to a contracted AI infrastructure model. The thesis now requires not just Phase I delivery completion but a clear quarter of material Data Centers revenue and positive adjusted segment EBITDA before the hybrid valuation can be justified. Until then, the investment case leans more heavily on crypto market recovery than on the AI data center catalyst.